fmicb-10-01169 May 28, 2019 Time: 14:28 # 1

ORIGINAL RESEARCH published: 28 May 2019 doi: 10.3389/fmicb.2019.01169

Metabarcoding Insights Into the Trophic Behavior and Identity of Intertidal Benthic Foraminifera

Panagiota-Myrsini Chronopoulou1*, Iines Salonen1*, Clare Bird2, Gert-Jan Reichart3 and Karoliina A. Koho1

1 Aquatic Biogeochemistry Research Unit, Ecosystems and Environment Research Programme, Faculty of Biological and Environmental Sciences, University of Helsinki, Helsinki, Finland, 2 Biological and Environmental Sciences, University of Stirling, Stirling, United Kingdom, 3 Department of Ocean Systems, NIOZ-Royal Netherlands Institute for Sea Research and Utrecht University, Den Burg, Netherlands

Foraminifera are ubiquitous marine protists with an important role in the benthic carbon cycle. However, morphological observations often fail to resolve their exact taxonomic placement and there is a lack of field studies on their particular trophic preferences. Here, we propose the application of metabarcoding as a tool for the elucidation of the in situ feeding behavior of benthic foraminifera, while also allowing the correct Edited by: taxonomic assignment of the feeder, using the V9 region of the 18S (small subunit; SSU) Ramiro Logares, Institute of Marine Sciences (ICM), rRNA gene. Living foraminiferal specimens were collected from two intertidal mudflats Spain of the Wadden Sea and DNA was extracted from foraminiferal individuals and from the Reviewed by: surrounding sediments. Molecular analysis allowed us to confirm that our foraminiferal Franck Lejzerowicz, University of California, San Diego, specimens belong to three genetic types: Ammonia sp. T6, Elphidium sp. S5 and United States Haynesina sp. S16. Foraminiferal intracellular eukaryote communities reflected to an Russel J. S. Orr, extent those of the surrounding sediments but at different relative abundances. Unlike University of Oslo, Norway sediment eukaryote communities, which were largely determined by the sampling site, *Correspondence: Panagiota-Myrsini Chronopoulou foraminiferal intracellular eukaryote communities were driven by foraminiferal species, [email protected] followed by sediment depth. Our data suggests that Ammonia sp. T6 can predate on Iines Salonen iines.salonen@helsinki.fi metazoan classes, whereas Elphidium sp. S5 and Haynesina sp. S16 are more likely to ingest diatoms. These observations, alongside the use of metabarcoding in similar Specialty section: ecological studies, significantly contribute to our overall understanding of the ecological This article was submitted to Aquatic Microbiology, roles of these protists in intertidal benthic environments and their position and function a section of the journal in the benthic food webs. Frontiers in Microbiology Keywords: metabarcoding, benthic foraminifera, trophic strategy, benthic food web, benthic microbial ecology, Received: 03 December 2018 molecular phylogeny Accepted: 07 May 2019 Published: 28 May 2019 Citation: INTRODUCTION Chronopoulou P-M, Salonen I, Bird C, Reichart G-J and Koho KA Benthic foraminifera are ubiquitous, single-celled protists. Due to their opportunistic character (2019) Metabarcoding Insights Into the Trophic Behavior and Identity (e.g., Moodley et al., 2000; Woulds et al., 2007), foraminifera can take advantage of their of Intertidal Benthic Foraminifera. environment very efficiently and they are able to thrive in a wide variety of marine environments. Front. Microbiol. 10:1169. Their ecology is complex, with some species harboring photosynthetically active symbionts or doi: 10.3389/fmicb.2019.01169 kleptoplasts (e.g., Hallock, 2000; LeKieffre et al., 2018; Schmidt et al., 2018) and other various

Frontiers in Microbiology| www.frontiersin.org 1 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 2

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

endobionts (e.g., Bernhard, 2003; Tsuchiya et al., 2015; Bernhard is well-represented in public databases, and it captures a large et al., 2018), of which some may be used in direct carbon transfer eukaryotic diversity including that of protists (Amaral-Zettler to host foraminifera (Tsuchiya et al., 2018). Foraminifera are et al., 2009; Behnke et al., 2011; Pawlowski et al., 2011). However, generally considered as heterotrophic organisms with multiple this hypervariable region has not yet been considered for the feeding strategies. Of these, carnivory and predation are well- taxonomic placement of benthic foraminifera. documented among planktonic foraminifera (Boltovskoy and Here, for the first time, to the best of our knowledge, we Wright, 1976; Bé et al., 1977), however, for benthic foraminifera target the V9 hypervariable region of the 18S rRNA gene we rely only on experimental observations, which suggest that within benthic foraminiferal cells. In addition, the foraminiferal some species may pray on nematodes or other metazoans intracellular eukaryote communities are compared to those of (Suhr et al., 2008; Dupuy et al., 2010). Instead, a number their surrounding sediments to gain insights into the relative of experimental studies suggest that phototrophs provide an distribution of foraminiferal food sources in the sediment. important source of organic carbon and nutrients to benthic Moreover, the observed intracellular eukaryote diversity is linked foraminifera (Moodley et al., 2000; Nomaki et al., 2005, 2006; to external factors (e.g., site, habitat depth in sediment, and total Larkin et al., 2014; Jeffreys et al., 2015; LeKieffre et al., 2017). sedimentary organic carbon and nitrogen content), as parameters Generally, however, there is a distinct lack of in situ evidence like organic carbon availability and sediment depth have been of species-specific feeding modes and ecological relationships shown to be important in structuring the intertidal foraminifera among benthic foraminifera and sediment micro- and meiofauna community (e.g., Thibault De Chanvalon et al., 2015; Mojtahid due to the difficulties of studying these processes in nature. et al., 2016). The overall aim of this study is to identify species- Understanding species-specific feeding behaviors is crucial to specific trophic preferences of benthic foraminifera, and, in unraveling the adaptability strategies of benthic foraminifera in parallel to unravel their taxonomic identity. their habitats, understanding the benthic food webs structure and addressing implications for the global marine benthic biogeochemical cycles. MATERIALS AND METHODS Metabarcoding may provide new insights into life strategies and in situ feeding modes of foraminifera and allow the Site Description and Sampling identification of potential species-specific preferences. This Two intertidal mudflat localities (Supplementary Figure 1) approach has been successfully applied to investigate the were sampled in November 2015 at the Dutch Wadden microbiome and potential feeding preferences of marine Sea: Mokbaai (M) characterized by relatively sandy sediment eukaryotes, such as copepods (Ray et al., 2016) and nematodes with the presence of polychaete worm burrows (>10 cm (Schuelke et al., 2018). Recently, 16S rRNA metabarcoding was depth), and de Cocksdorp (C) characterized by non-burrowed also used to study the intracellular bacterial composition of clay/mud sediment. pelagic foraminifera to elucidate their ecological strategies (Bird One sediment core (10 cm internal diameter) per site was et al., 2017, 2018). Cloning and shallow Sanger sequencing sampled manually by pushing a core tube into the sediment have been recently used to demonstrate the multiple diatom during low tide and processed as described in Koho et al., associations within an individual benthic foraminifer, suggesting 2018; see detailed steps in Supplementary Figure 1C. In short, that the host can shuffle its symbionts in response to thermal three sub-cores (50 ml truncated syringes) were taken from stress (Schmidt et al., 2018). However, the application of the main core. Two of the sub-cores were transferred in a metabarcoding in benthic foraminifera is yet to be tested. nitrogen-filled glove bag and sliced with 1 cm intervals down to A good taxonomic resolution is essential in solving species- 10 cm depth. Porewater was removed, centrifuging the sediment, specific feeding preferences and potential niche and resource and the solid phase was frozen to –20◦C and transferred to partitioning among foraminiferal population. For planktonic the University of Helsinki, where it was freeze-dried. Then, foraminifera, cryptic species have been shown to display niche sedimentary organic carbon and total nitrogen was measured differentiation within the water column (Weiner et al., 2012) with a Leico TruSpec R Micro, following homogenization and as well as geographically on a spatial scale (Aurahs et al., decalcification (1 M HCl). The third sub-core was also sliced 2009). Metabarcoding allows not only the identification of prey at 1 cm intervals down to 10 cm sediment depth and used but also the cryptic diversity of the feeder that is not readily to obtain environmental DNA (eDNA; referred to as sediment distinguished morphologically (e.g., Miller et al., 1982; Pawlowski DNA) samples and foraminiferal specimens. Each sediment slice and Holzmann, 2008; Schweizer et al., 2011; Pillet et al., 2012; was subsampled (ca. 1–1.5 g sediment) with a sterile plastic Darling et al., 2016; Roberts et al., 2016; Lei et al., 2017). The 37f spatula, the subsample was immediately frozen in liquid nitrogen hypervariable region of the 18S (SSU) rRNA gene is commonly and kept stored in –20◦C until eDNA extraction. The rest of used in foraminiferal molecular studies (Pawlowski, 2000). As the slice was sieved with filtered seawater through a 125 µm this helix region is foraminifera-specific and able to identify mesh and intact foraminiferal cells with visible protoplasm foraminifera to species level (Lecroq et al., 2011), it has been picked under a microscope (see Supplementary Table 1 for proposed as a DNA barcode (Pawlowski and Holzmann, 2014). details on collected living specimens). Vitality was confirmed Yet, the 37f region wider use in foraminiferal identification is based on movement of foraminifera under oxygenated conditions impeded by the under-representation in public databases. In (see Koho et al., 2011), and foraminifera specimens were contrast, the V9 hypervariable region of the 18S rRNA gene identified to genus level morphologically. Subsequently, each

Frontiers in Microbiology| www.frontiersin.org 2 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 3

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

living specimen was washed three times with sterile artificial Sequencing and Genomics at the Institute of Biotechnology, seawater, transferred into RNAlater solution (InvitrogenTM), Helsinki Institute of Life Science (HiLIFE). which dissolves the calcite test, and stored at +4◦C until further molecular analyses. Processing of Sequences and Phylogenetic Analysis DNA Extraction, Amplification, and Raw reads were de-multiplexed to samples based on their Sequencing barcode sequences and MiSeq overhangs, primers, and barcode DNA was extracted from foraminiferal individuals following sequences were removed as described in Salava et al.(2017). the DOC (sodium deoxycholate) method (Holzmann and Sequences were assembled to paired-end reads and quality- Pawlowski, 1996). Before placement in the DOC buffer, the filtered in Mothur version 1.39.5 (Schloss et al., 2009). Minimum naked foraminiferal cells were washed again 3–5 times in sterile and maximum sequence lengths were set to 122 and 151 bp, artificial seawater (Red Sea’s Coral Pro Salt, salinity adjusted to respectively. No ambiguous sequences were allowed and the 29 ), to clean the cells of any surficial organisms and eliminate maximum number of homopolymers was set to 8. Quality-filtered RNAhlater traces (see Bird et al., 2017). The partial SSU rRNA reads were aligned against the SILVA database (release 128) and gene (approximately 550 base pairs (bp)) of two specimens (M1C chimeric sequences were removed with the implementation of and M5B) was genotyped by conventional methods according to UCHIME algorithm (Edgar et al., 2011) in Mothur. Taxonomic (Darling et al., 2016). Sediment DNA (ca. 0.25 g) was extracted assignment of all sequences was performed in Mothur against using the PowerSoil R DNA Isolation Kit (MoBio, Carlsbad, CA, the SILVA database and taxonomic information was used in United States), according to the manufacturer’s instructions. downstream clustering. Clustering into Operational Taxonomic DNA from foraminifera and sediment samples was amplified Units (OTUs) was done using an arbitrary chosen 95% alongside three extraction controls containing no template with similarity sequence cutoff (e.g., Caron et al., 2009) in order either (i) DOC and artificial seawater (two replicates) and (ii) the to aggregate variation due to sequencing and PCR errors. buffers of MoBio PowerSoil R DNA Isolation Kit. In addition, non- Consensus taxonomy for each OTU was determined at 0.05 template PCR controls of the first and second (indexing) PCR (see distance level. OTUs assigned to Foraminifera by SILVA were below) were sequenced. further compared to the PR2 (version 4.7) database (Guillou The V9 region of the 18S rRNA gene was targeted with et al., 2013) to achieve genus level assignment. Representative the 1389F/1510R primers described by Amaral-Zettler et al. sequences for each OTU were determined in Mothur as the (2009), and widely used in ecological studies for the investigation centroids (sequence with the smallest distance to the other of eukaryotic diversity (e.g., de Vargas et al., 2015; Sawaya sequences) of the distance matrix created at the clustering et al., 2019; Pitsch et al., 2019). Primers were modified at the stage. The representative sequences of OTUs that remained 50 end to include overhang sequences (Illumina adapters) unclassified with the SILVA database, were aligned in a stand- for the downstream sequencing (forward overhang (37 bp): alone BLAST search (Altschul et al., 1990) against the NCBI’s 50-ATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT- non-redundant nucleotide database. BLAST results were also 30; reverse overhang (34 bp): 50- used to confirm the identity of foraminiferal specimens at the GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT-30). genus level (Table 1). Amplification reactions were performed on an Applied OTUs with ≤ 8 and ≤ 10 sequence reads across the Biosystems Veriti 96 Well Thermal Cycler, using the Phusion foraminiferal and sediment datasets, respectively, were removed. Mastermix (Thermo Fisher Scientific) and following the We set these thresholds empirically based on the cumulative manufacturer’s protocol. PCR conditions for foraminiferal DNA sum of OTUs removed at increasing threshold in order to were as follows: 98◦C for 1 min, 25 cycles of 98◦C for 10 s, 67◦C reduce the amount of rare diversity while preserving our for 15 s and 72◦C for 15 s, 12 cycles of 98◦C for 10 s, 72◦C for sequencing effort (see Supplementary Figure 2). Filtering 15 s and 72◦C for 30 s, with a final elongation of 72◦C for 1 min. retained 99.86 and 99.03% of the total reads count for the PCR conditions for sediment DNA were the same, except for the foraminiferal and sediment dataset, respectively. Only two annealing temperature (72◦C) and cycle numbers (25–30 cycles). OTUs (unclassified Eukaryota) were excluded from the sediment Duplicate PCRs were performed and pooled in equal volumes, to dataset, as due to their abundance in the non-template PCR minimize the intra-sample variance and obtain enough amplicon control (39,668 and 6,759 sequences, accounting for 84.15 volume for Illumina library preparations. Pooled samples, and 14.34% of reads in the non-template PCR controls but including negative controls, were quality-checked on 1.5% w/v only 0.46 and 0.37% of reads on average in the samples) agarose gels. Prior to sequencing, PCR products were purified they were considered contaminants in the PCR reactions. and a second indexing PCR (P7 unique index attached) was One more OTU was excluded because it was abundant in performed followed by magnetic bead purification as described the kit extraction control (137 sequences, accounting for in Salava et al.(2017). In order to mitigate the possibility of 29.40% of reads in the control but 0.00007% on average in cross-contamination due to mistagging (Esling et al., 2015), the samples) indicating that it is a contaminant of the kit unique barcodes were selected for the indexing PCR using reagents. DOC extraction buffer controls returned low numbers BARCOSEL (Somervuo et al., 2018). Samples were sequenced of sequences (half the average number of sequences in the on the Illumina MiSeq platform of the Laboratory of DNA samples), which could either not be aligned to SILVA’s 18S

Frontiers in Microbiology| www.frontiersin.org 3 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 4

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

TABLE 1 | Foraminiferal specimens and their identity.

Specimen Depth (cm) ID (PR2) Genotype Closest relative to most BLAST No. of foram % reads in most code abundant OTU (BLAST) ID (%) OTUs abundant OTU

M1B 0–1 Amm NA A. aomoriensis (GQ853573) 100 18 80.46 M1C∗ 0–1 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 23 99.52 M1D 0–1 Amm NA A. aomoriensis (GQ853573) 100 8 48.62 M2B 1–2 Amm NA A. aomoriensis (GQ853573) 100 21 81.37 M2E 1–2 NA NA NA NA 7 80.39 M3A 2–3 Amm NA A. aomoriensis (GQ853573) 100 19 77.44 M3B 2–3 Amm NA A. aomoriensis (GQ853573) 100 19 78.39 M3D 2–3 Hay Hay sp. S16 Haynesina sp. S16 (KX962996) 99 28 95.95 M4C 3–4 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 17 98.95 M4D 3–4 NA NA NA 100 5 63.64 M5B∗ 4–5 Amm Amm sp. T6 A. aomoriensis (GQ853573) 100 19 94.73 M6A 5–6 Amm NA A. aomoriensis (KT989509) 100 7 84.77 M6B 5–6 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 5 96.10 M7A 6–7 NA NA NA NA 5 80.31 M7D 6–7 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 26 99.52 M8A 7–8 Amm NA A. aomoriensis (GQ853573) 100 17 65.17 M8B 7–8 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 27 78.10 M8D 7–8 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 16 99.69 M9B 8–9 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 22 99.68 M9F 8–9 Amm NA A. aomoriensis (GQ853573) 100 9 70.50 M10B 9–10 Hay Hay sp. S16 Haynesina sp. S16 (KX962996) 99 25 89.90 M10C 9–10 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 20 99.64 M10D 9–10 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 20 99.75 C1A 0–1 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 19 99.68 C1B 0–1 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 24 99.65 C2D 1–2 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 25 95.9 3 C3B 2–3 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 24 84.88 C4C 3–4 Elph Elph sp. S5 Elphidium sp. S5 (KX962814) 100 23 99.48

∗For specimens M1C and M5B genotyping was conducted as described in Darling et al.(2016). “M” indicates foraminiferal specimens from Mokbaai and “C” from de Cocksdorp. “Amm” stands for Ammonia sp., “Elph” for Elphidium sp. and “Hay” for Haynesina sp. NA = not applicable. Molecular identification at genus level was done via taxonomic assignment of the obtained foraminiferal OTUs based on the Protist Ribosomal Reference database (PR2) in Mothur. Samples M2E, M4D and M7A could not be assigned taxonomy at genus level, due to very low abundance of foraminiferal retrieved sequences from these specimens, thus their microscopic identification was used (Ammonia sp. for M2E and M7A; Elphidium sp. for M4D). Genetic types are as described in Darling et al.(2016), and here they refer to the closest relatives of the foraminiferal OTUs in each specimen based on a BLAST search. For comparison, the BLAST result for the most abundant foraminiferal OTU in each specimen is also given.

database or were assigned to prokaryotes and thus filtered sites and assuming a certain fraction of sites (15.32%) to out by the Mothur pipeline with no interference to the be evolutionarily invariable) according to BIC (Bayesian downstream analysis. Information Criterion) (Hall, 2013). The tree was edited in In order to compare the diversity of eukaryotic communities Dendroscope (version 3.5.9; Huson et al., 2007) and Adobe found in foraminiferal hosts and in the surrounding Illustrator CC (2014 release). sediment, OTUs belonging to phylum Retaria (called TF = Texel Foraminifera) were excluded from both datasets. Statistical Analysis Sediment OTUs are hereafter called “TS” (standing for Statistical analysis was done in R (version 3.4.2), using the Texel sediments) and intracellular foraminiferal eukaryote packages phyloseq (version 1.22.3) (McMurdie and Holmes, OTUs called “TIFC” (standing for Texel intracellular 2013) and vegan (version 2.4-4) (Oksanen et al., 2015). DCA foraminiferal content). (detrended correspondence analysis) indicated that both the Representative sequences of all the TFs and their closest foraminiferal and sediment datasets are heterogeneous (length of relatives were aligned using the muscle algorithm (v3.8.31, first DCA axis > 4 standard deviations), thus unimodal models Edgar, 2004) and edited in MEGA7 (Kumar et al., 2016). were applied for multivariate analysis. Available environmental Maximum likelihood (ML) phylogenetic tree was constructed data [sedimentary organic carbon and total nitrogen contents using MEGA7, after performing a “best model” analysis and their molar ratio (C/N)], sampling site and sample depth to select the best substitution model (Kimura 2-parameter range (0–2, 2–6, and 6–10 cm) were considered as potential model with discrete Gamma distribution rates among explanatory variables for the observed community variance.

Frontiers in Microbiology| www.frontiersin.org 4 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 5

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

Automatic stepwise model building (ordistep in package vegan) rest of the foraminiferal samples (see Figures 2, 4A), thus they was applied, in order to select the best fitting model based on were included in subsequent analysis. the Akaike information criterion (AIC) and using permutation tests. Multicollinearity was checked by calculation of the Identification of Foraminiferal variance inflation factors (VIFs) and only factors with VIF < 5 Specimens and Phylogenetic Analysis of were considered. Foraminiferal OTUs Taxonomic identification was based on the TF with the greatest Accession Numbers number of reads in each specimen (87.22% ± 13.70% average The DNA sequences representative of OTUs reported in this foraminiferal reads across specimens; see last column of Table 1). study were deposited in the Genbank database. A total of 65 Specimens M2E, M4D, and M7A could not be assigned to foraminiferal sequences (TF) are under the accession numbers genus level, so their microscopic identification was adopted. All MK011309 – MK011373, 445 foraminiferal intracellular our specimens fall within the order Rotaliida. In Mokbaai, 11 content sequences (TIFC) under the accession numbers specimens were identified as Ammonia sp., 10 as Elphidium sp. MK012677 – MK013121, and 1,571 sediment sequences and 2 as Haynesina sp., whereas all 5 specimens of de Cocksdorp (TS) under the accession numbers MK020770 – MK022340. were identified as Elphidium sp. (Table 1). Moreover, the raw fastq files were deposited to SRA under For the maximum likelihood tree, representative TF sequences the Sequence Read Archive (SRA) BioProject accession were aligned (ca. 117 bp; positions 1389–1510 of 18S rRNA number PRJNA472012. gene) alongside 11 sequences of their closest relatives (97–100% similarity) and 37 sequences of known foraminiferal species. The majority of TF OTUs (21 TF, corresponding to 64.83% of all RESULTS foraminiferal sequences) are similar (≥99% BLAST similarity) to Elphidium genetic type S5 and form a large clade (81% ML Taxa (OTUs) Obtained and Sequencing bootstrap support), including also genetic types S3, S4, and S13 Depth (Figure 1). Another big cluster on the tree, with 86% bootstrap support, is that of Ammonia sp., comprising the genetic types DNA was analyzed from within 23 foraminiferal specimens T6, T3V, and T3S (A. batava). The second most abundant group from Mokbaai and 5 specimens from de Cocksdorp (Table 1). of our sequences (16 TF; 24.99% of all foraminiferal sequences; Additionally, sediment samples obtained from the same depths 97-100% BLAST similarity to Ammonia aomoriensis (GQ853573) as foraminiferal specimens (0–10 cm for site M, 0–4 cm for site and >99% to Ammonia sp. T6 (KT989509)) falls within this C) were used for metabarcoding along with the foraminifera. cluster. Finally, there is a cluster of Haynesina sp.-related OTUs A total of 2,847,274 sediment and 5,227,694 intracellular (25 TF), which is not a well-supported clade (only 20% bootstrap foraminiferal sequence reads were obtained, which after quality support). Among this cluster 16 TF (6.10% of all foraminiferal filtering were reduced to 1,881,013 for the sediment and 3,654,067 sequences) are highly similar (>98%) to Haynesina sp. S16 for the foraminiferal dataset. Chimera check removed another (KX962996, KX962992). 0.56% of the sediment and 0.13% of the intracellular foraminiferal reads. The remaining reads were clustered into 6,949 OTUs for Foraminiferal Intracellular Eukaryote the sediment and 3,011 OTUs for the intracellular foraminiferal dataset. After filtering out OTUs with low number of reads (see Content Compared With Surrounding Materials and Methods and Supplementary Figure 2) and non- Sediment Eukaryote Communities eukaryote OTUs, 1,608 OTUs were obtained from the sediment TIFC reflected TS, but clear differences were observed in relative and 510 OTUs from the foraminiferal dataset, of which 65 OTUs abundances (Figure 2). For example, diatoms (class Diatomea in (TF) were assigned to phylum Retaria and all other 445 OTUs to Figure 2) were the most abundant eukaryotes in the majority their intracellular eukaryote content (TIFC). After the exclusion of the foraminiferal specimens (51.36% relative abundance on of Retaria OTUs from the sediment data, 1,571 OTUs (TS) average). They were also common in sediments, but generally remained for further analysis. at lower relative abundances (22.67% relative abundance on Rarefaction analysis indicates that the filtered OTU dataset average). Alpha diversity measured using either the Shannon reaches asymptote levels, allowing for richness comparison or Simpson index was significantly higher for TS than TIFC among samples for both the sediment and intracellular (ANOVA, p < 0.001, Figure 3). foraminiferal datasets (Supplementary Figure 3). One sample The composition of TIFC appeared to be species-specific that exhibits the same OTU richness as the controls and is (Figure 2). The intracellular community of the two Haynesina distant from the rest of sediment samples was discarded from sp. specimens consisted entirely of diatoms, and the same was the TS dataset (C1, Supplementary Figure 4). In TIFC dataset, true for two Elphidium sp. specimens (M7D, M10C). A variety most of the samples reached a satisfactory sequencing depth (7 of diatom genera was found in all three species (Supplementary samples above the upper quartile (127,312 reads per sample) Figure 5). Pennate genera, such as Climacosphenia sp. and and 14 samples above the median (90,673 reads per sample, Petrodictyon sp. were common in Elphidium sp. of surface Supplementary Figure 3B). Samples with less reads (e.g., M4C, sediments, whereas Elphidium sp. specimens from deeper M4D, M7A, M6B) had similar composition and grouped with the sediments contained more Thalassiosira sp. and genera

Frontiers in Microbiology| www.frontiersin.org 5 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 6

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

FIGURE 1 | Maximum likelihood phylogenetic tree of the foraminiferal OTUs and their closest relatives. The tree was built based on partial SSU rDNA sequences (about 117 bp) and inferred using the ML method with the Kimura 2-parameter model. Collapsed branches are indicated by a triangle/polygon. The tree was rooted on Allogromia sp. (X86093). Bootstrap support values over 1000 replicates are shown at the nodes. The number in parenthesis following the TF sequences indicates their % relative abundance over the total number of foraminiferal sequences. The bar represents 0.1 average nucleotide substitutions per site.

Frontiers in Microbiology| www.frontiersin.org 6 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 7

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

FIGURE 2 | Relative abundance of eukaryote taxa at class level for foraminiferal intracellular eukaryote content (showing classes with > 2% abundance, i.e., 90.02% of all reads; foraminiferal OTUs excluded from the analyses) and communities of the surrounding sediments (showing classes with > 0.5% abundance, i.e., 83.20% of all reads). Foraminiferal species (Ammonia sp., Elphidium sp., Haynesina sp.) and sampling sites (de Cocksdorp, Mokbaai) are shown on the top grid. Taxa that are similar to uncultured eukaryotes are indicated by “uncult” followed by information on the environment of their closest relatives.

of the family Mediophyceae. Alongside diatoms, some Elphidium sp. specimens) and Acoela (e.g., 20% in M2B; none in Elphidium sp. specimens contained dinoflagellates (e.g., Elphidium sp. specimens). class Dinophyceae, 13–31% relative abundance in M4C, M6B, Non-metric multidimensional scaling (nMDS) analysis of M9B, M10D), ciliates (class Intramacronucleata, 23–32% TIFC (Figure 4A) showed that the three foraminiferal species are relative abundance in C1A, C3B, M4C) and fungal groups well separated in the ordination space, followed by separation (e.g., class Saccharomycetes 39 % relative abundance in M1C, based on the depth range from which the specimens derived. and class Exobasidiomycetes 51% in C3B and 52% in M4D). TIFC of Ammonia specimens generally clustered together, Metazoan classes were generally more abundant in Ammonia however three specimens (M2E, M3B, M2B) were separated sp. specimens, i.e., Maxillopoda (relative abundance 10% in from the rest and closer to Elphidium and Haynesina specimens. M9F to 76% in M5B; only 3–22% in some Elphidium sp. TIFC in these specimens was dominated by diatoms, as was specimens), Nematoda (e.g., the class Chromadorea with the case with Elphidium and Haynesina specimens. Species 95% in M1D, 18% in M8A, 49% in M9F, but only 1–6% in was a significant factor (PERMANOVA, F = 2.884, p = 0.001)

Frontiers in Microbiology| www.frontiersin.org 7 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 8

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

FIGURE 3 | Summary of the alpha diversity, calculated by (A) Shannon and (B) Simpson indices, of foraminiferal intracellular content (excluding foraminiferal OTUs) and sediment communities. Foraminiferal communities were grouped per depth interval. There are multiple foraminiferal specimens for each depth interval (see Table 1, here shown by boxplots) but always one sediment sample per depth interval. Boxplots show the median (middle line) diversity; the lower and upper hinges correspond to the first and third quartiles (the 25th and 75th percentiles) of the diversity range; the upper and lower whiskers extend from the hinge to the largest and lowest value no further than 1.5 ∗ IQR from the hinge (where IQR is the inter-quartile range, or distance between the first and third quartiles).

for the observed community variance, followed by sediment range, pairwise comparisons carried out separately for each depth range (PERMANOVA, F = 1.447, p = 0.040). This was species within each depth range (0–2, 2–6, and 6–10 cm), also true for the distribution of intracellular diatom genera indicated no significant differences among TIFC of the same (species: PERMANOVA, F = 2.030, p = 0.016; depth range: depth range groups (pairwise MANOVA, p > 0.14 within PERMANOVA, F = 1.530, p = 0.047). In contrast to overall and among species, with Benjamini–Hochberg adjustment). TIFC community composition, which was driven by the depth Additionally, the significance of site (de Cocksdorp vs. Mokbaai

Frontiers in Microbiology| www.frontiersin.org 8 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 9

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

FIGURE 4 | Non-metric multidimensional scaling (nMDS) plots of (A) foraminiferal intracellular eukaryote content (excluding foraminiferal OTUs) and (B) communities of the surrounding sediments. Samples from different sediment depths (cm) are grouped in three depth ranges: 0–2, 2–6, and 6–10 cm. “M” indicates foraminiferal specimens and sediment samples from Mokbaai; “C” indicates foraminiferal specimens and sediment samples from de Cocksdorp. nMDS was based on a Bray-Curtis distance and the stress for foraminifera was 0.2243, whereas for sediments 0.1280.

specimens) was evaluated, after excluding Mokbaai specimens range (ANOVA, F = 1.676, p = 0.004). All the other factors from 5 cm and deeper, as no living specimens were found (including organic carbon and nitrogen contents) were not deeper than 4 cm depth in de Cocksdorp. The analysis significant but contributed to the overall variance explained by showed that site was not a significant factor (PERMANOVA, the constraints of the model (48.28%). F = 1.038, p = 0.401). In contrast to the foraminiferal intracellular eukaryote content, the sediment eukaryote community between Mokbaai and de Cocksdorp was different (Figure 4B). Site DISCUSSION was the most significant factor in sediments (PERMANOVA, F = 3.658, p = 0.001), followed by depth range (PERMANOVA, Metabarcoding of the 18S V9 Region: A F = 2.056, p = 0.009). Subsequently, Canonical Correspondence Analysis Useful Tool for the Taxonomic Placement (CCA) was performed to account for the impact of various of Intertidal Foraminifera environmental factors on the observed foraminiferal intracellular Correct taxonomy is pivotal in understanding species-specific eukaryote content variance (Figure 5). A total of 24.80% of the trophic behavior and benthic food-web structure. Based on this observed community variance was explained by the constraints study, metabarcoding of the 18S V9 region and using PR2 (foraminiferal species, sediment depth range and the per-depth (Guillou et al., 2013) as reference database allows determining the range average nitrogen (N) and organic carbon (C), as well taxonomic placement of foraminiferal specimens. The taxonomy as their ratio (C/N); see Supplementary Table 2 for C and N suggested by PR2 was confirmed by BLAST results (Table 1) concentrations). Overall, our chosen CCA model was significant and further supported by phylogenetic analysis (Figure 1). (ANOVA, F = 1.154, p = 0.03). Foraminiferal species was the TF OTUs were assumed to derive from the specimens’ own main driving factor in explaining the foraminiferal intracellular DNA. We cannot preclude the possibility of foraminifera eukaryote content (ANOVA, F = 1.421, p = 0.004), followed by praying on other foraminifera (e.g., Lipps, 1983), however, sediment depth range (ANOVA, F = 1.160, p = 0.041). No other on average 87% of the TF reads within our specimens were factor contributed significantly to the observed foraminiferal taxonomically assigned (and confirmed by phylogenetic analysis) intracellular eukaryote content variance. A similar CCA model to the same foraminiferal species as the species assigned based was built for the sediment communities (Supplementary on morphology. Thus, in this case, foraminifera cannibalism is Figure 6), which was overall significant (ANOVA, F = 1.867, unlikely to play an important role. Morphological identification p = 0.004) and confirmed that site was the most significant factor of some foraminiferal specimens is a difficult task and can (ANOVA, F = 2.566, p = 0.001), followed by sediment depth lead to wrong taxonomic assignment. For example, similar

Frontiers in Microbiology| www.frontiersin.org 9 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 10

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

FIGURE 5 | Canonical Correspondence Analysis (CCA) of foraminiferal intracellular eukaryote content (excluding foraminiferal OTUs) and potential explanatory variables. Specimens from different sediment depths (cm) are grouped in three depth ranges: 0–2, 2–6, and 6–10 cm. “M” indicates foraminiferal specimens from Mokbaai and “C” from de Cocksdorp. Arrows, indicating the correlation between the canonical axes and the explanatory variables, are only shown for the significant variables. Average organic carbon content (in weight % of dry sediment), average total nitrogen content (in weight % of dry sediment) and average C/N per depth range (C mol/ N mol) were also included in the CCA model but were not significant (p > 0.1). Organic carbon and nitrogen content values are shown in Supplementary Table 2.

morphologies have been documented for different Ammonia sp. region is a very small region of the 18S rRNA gene. In this case, genetic types, such as T1, T2, T6, and T10 (Hayward et al., 2004; the alignment length was only about 117 nucleotide sites, which, Schweizer et al., 2011). The same is also true for Elphidiidae in addition to the genetic variability within elphidiids, constrains (e.g., Pawlowski and Holzmann, 2008; Darling et al., 2016), the robustness of our phylogenetic analysis. The observed low particularly in the case of small specimen sizes. Thus, the bootsrap support values make the phylogenetic relationships importance of integrating morphological and molecular results to difficult to intrepret, and, hence the phylogentetic tree here serves secure identification and taxonomic placement of foraminiferal only as as a visualization tool for within-clade sequence similarity. species has been recognized and established in recent benthic Comparison of our sequences to databases is sufficient for a foraminiferal studies (e.g., Schweizer et al., 2008; Pillet et al., 2013; secure taxonomic assignment (similarities ≥ 97%). Darling et al., 2016; Roberts et al., 2016). However, care should Phylogenetic analysis confirms that our specimens are part be taken when assigning taxonomy at genus/species level, as of the order Rotaliida, belonging to Elphidiidae, Rotaliidae, and results may differ depending on the database used. For example, Nonionidae families (Holzmann and Pawlowski, 2017). The large based on our results, SILVA database tends to assign sequences Elphidium-related clade on our tree (81% ML bootstrap support, of the order Rotaliida to Ammonia sp., although BLAST and Figure 1) is matching clade F of the phylogenies presented in phylogenetic analysis confirmed that many of our specimens Pillet et al.(2013) and Darling et al.(2016). The morphologically belonged to Elphidium sp. or Haynesina sp. The PR2 database similar but distinct genetic types S4 and S5 is a good example of was superior in the assignment of our benthic foraminiferal the taxonomic confusion within elphidiids (Roberts et al., 2016), sequences and it has also been curated to include all planktonic as genetic type S4 has been considered as part or subspecies of foraminiferal rDNA sequences (Morard et al., 2015, 2018). We E. excavatum, till the latest suggestion by Darling et al.(2016) therefore recommend the use of the PR2 database for the to assign the name E. clavatum to the genetic type S4 and the assignment of rotaliid foraminifera at genus level, yet we stress name E. selseyense to genetic type S5. Elphidium sp. genetic type the importance of following up with phylogenetic analysis for S5 has been found before in the Mokbaai mudflat (Schweizer secure identification. Nonetheless, care should be taken as the V9 et al., 2011; Jauffrais et al., 2018) and in other mudflats in the

Frontiers in Microbiology| www.frontiersin.org 10 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 11

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

United Kingdom (Schweizer et al., 2011; Darling et al., 2016) and of representative genetic types in the same clades A-F (except France (Ertan et al., 2004), and there have also been occurrences members of clades E and D that cluster in sister branches rather in the Baltic Sea (Schweizer et al., 2011). It seems to be a rather than in the same one on our tree). Even though a thourough widespread intertidal taxon, tolerant to relatively large variations phylogenetic placement of the various elphidiid genetic types is of temperature and salinity (Darling et al., 2016). The rest of outside the scope of this study, our results are consistent with the the Elphidium sp. in our phylogeny form separate clades, which established clades of the aformentioned studies. Notably, clade indicates a paraphyletic group and is in agreement with previous F in our analysis branches separately from clades A and B-E, phylogenetic placements (Schweizer et al., 2011; Pillet et al., which matches better the second scenario presented in Pillet et al. 2013; Darling et al., 2016). For example, clade A of Pillet et al. (2013). According to this scenario, rooting is done on Ammonia (2013) and Darling et al.(2016) with E. williamsoni (genetic type sp. and clade F branches separately from the rest of the clades, S1), E. macellum [Patagonia branch on Darling et al.(2016)], suggesting a closer evolutionary relationship between elphidiids E. margaritaceum 1 (genetic type S9) and E. aculeatum (genetic and nonioiniids. type S10), is a separate branch on our phylogeny, which clusters Our phylogenetic analysis corroborates the BLAST and PR2 together with a Rhizaria sequence retrieved from the waters of the results for the assignment of the genetic types Elphidium sp. Scotian Shelf (Dasilva et al., 2014). S5, Haynesina sp. S16, and Ammonia sp. T6 to our specimens, We only had two Haynesina sp. specimens, both retrieved which is consistent with the biogeographic distribution of these from the Mokbaai mudflat. All the Haynesina-related OTUs were genetic types. Moreover, the molecular identification is supported similar (>98% BLAST similarity) to genetic type S16 (Haynesina by SEM observations (see Figure 2 for Haynesina sp. S16 and germanica), forming part of clade C (Pillet et al., 2013; Darling Figure 5 for Elphidium sp. S5 in Jauffrais et al., 2018; Figure 1 for et al., 2016). However, the bootstrap support for this clade on our Ammonia sp. T6 in Koho et al., 2018) of specimens sampled from tree is extremely low (20%, Figure 1). This group of sequences is the same sites, as these match the morphological characteristics branching with H.orbiculare, which alongside S16 is part of clade of the above genetic types. C in Pillet et al.(2013) and Darling et al.(2016). In addition, E. asklundi appears on this branch in our phylogeny, whereas it is part of the sister clade D in the aforementioned studies. Trophic Preferences of Intertidal Haynesina sp. S16 has been retrieved from sediments in Den Foraminifera Oever and Texel, Netherlands (Schweizer et al., 2008), and it Here, for the first time, we used a metabarcoding approach has a similar geographic distribution to that of Elphidium sp. S5 to investigate in situ feeding patterns of intertidal benthic (Darling et al., 2016). foraminifera. This method, although with some known pitfalls Ammonia sp. sequences form a separate clade (86% ML related to amplification biases (e.g., Logares et al., 2014; support, Figure 1) on our phylogenetic tree, consisting of two Pawluczyk et al., 2015), is known to perform better compared branches. The first branch is that of Ammonia genetic types T2A to conventional amplicon sequencing, as it allows an in-depth (A. aberdoveyensis) and T2B, recently suggested as subgroups of community investigation. Our results successfully show the T2, based on both SSU and LSU (large subunit) rDNA, by Bird distinct food preferences of different foraminiferal species despite et al.(2019). The second one is that of genetic types T6 (often them inhabiting the same benthic environment. If foraminifera called A. aomoriensis), T3S (A. batava) and T3V. Our Ammonia were randomly deposit feeding on sediments and ambient sequences were similar (>97%) to a specimen from the Kiel Fjord eukaryotes, their intracellular eukaryote communities would be (SW Baltic Sea), identified as A. aomoriensis (GQ853573). The expected to be (i) similar between species and (ii) a close second Ammonia branch (63% ML support, Figure 1) on our reflection of the sediment composition. This was not the case phylogenetic tree is in agreement with the results based on partial as the constrained multivariate analysis (Figure 5) indicates SSU and LSU sequences (Schweizer et al., 2011), where Ammonia that foraminiferal species is the driving factor in shaping TIFC. sp. specimens from the Kiel Fjord cluster with the genetic type T6. In addition, whilst the sediment community was significantly This cosmopolitan genetic type has been found across different different at the two study locations (Supplementary Figure 6), geographic areas, e.g., in the North Sea (Langer and Leppig, the TIFC was not affected by site. Furthermore, the greater alpha 2000), in the sediments of brackish waters of Japan (Nomura and diversity of the TS compared to TIFC (Figure 3) suggests that Seto, 1992, Nomura, 2003; Takata et al., 2006) and in the Yellow foraminifera may have some preferences with regards to what Sea of China (Xiang et al., 2008). The figure holotype of T6 from taxa they feed on from their environment and therefore do Honshu, Japan was named A. aomoriensis (Asano, 1951) but its not simply reflect the biota in the surrounding sediments. This adoption for genetic type T6 is under debate (Hayward et al., diversity, however, can only be regarded as a proxy of potential 2004; Bird et al., 2019). trophic preferences and not as solid evidence, as the difference in In previous studies the number of nucleotide sites used in sample material (1–1.5 g sediment vs. a single foraminiferal cell) phylogenies was considerably larger [1686 sites in Pillet et al. could have an effect on the observed alpha diversity. (2013) and 601 sites in Darling et al.(2016)] than ours (117 In most of our Ammonia sp. specimens, targeting the 18S sites), therefore producing statisticaly more robust topologies. V9 region revealed an enrichment of metazoan classes (e.g., However, there is generally a good agreement between published Acoela, Chromadorea, Maxillopoda), implying that in addition tree topologies [Figure 1 in Pillet et al.(2013), Figure 2 in to feeding on phototrophs (e.g., diatoms), Ammonia sp. has a Darling et al.(2016)] and ours ( Figure 1), with the placement tendency toward active predatory behavior. Indeed, Ammonia

Frontiers in Microbiology| www.frontiersin.org 11 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 12

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

tepida has been shown in laboratory experiments to actively 2018). Our data confirms that Elphidium sp. and Haynesina sp. entrap nematodes with its pseudopodial network and empty the contain a wide range of diatoms (Supplementary Figure 5), nematode’s soft tissue within 18 h of initial contact (Dupuy et al., thus implying that the kleptoplasts may have originated from 2010). In addition, a few other benthic foraminifera have been a variety of diatom species. In addition, our data shows that shown to feed on metazoans (Langer and Bell, 1995; Goldstein, the foraminiferal intracellular diatom community changes with 1999; Suhr et al., 2008). Until now, however, in situ evidence sediment depth. As photosynthesis is restricted to surface of this behavior is lacking. Further in situ observations on sediments, where light is readily available, our observations different benthic foraminiferal species are needed to elucidate suggest that in the surface, pennate diatoms found inside their carnivorous behavior in different environmental conditions Elphidium sp. specimens may be linked to kleptoplasty, and and to fully understand the position of foraminifera in the diatoms found in specimens from deeper sediments (e.g., benthic food web. Thalassiosira sp.) may be taken up predominantly as a food The intracellular eukaryote communities of our Elphidium sp. source. However, 16S rRNA gene metabarcoding of more and Haynesina sp. specimens were mainly dominated by diatoms. specimens is needed to confirm diatom specificity patterns in the Foraminiferal ingestion of diatoms has been documented in intracellular foraminiferal communities. numerous feeding experiments (e.g., Larkin et al., 2014; Jeffreys The intracellular eukaryotic community of some of our et al., 2015; LeKieffre et al., 2017). In addition, Austin et al. Elphidium sp. specimens also contained a high relative abundance (2005) observed that H. germanica specimens were drawing the of dinoflagellates and ciliates. In the feeding study of Lee provided diatoms toward their aperture with their pseudopodia et al.(1966), various species of littoral foraminifera, including and SEM images indicated a characteristic cracking pattern of Elphidium sp., were introduced to multiple carbon sources, the diatom frustules. In another laboratory experiment where including dinoflagellates. No dinoflagellates were ingested, and H. germanica was provided with diatoms and sewage-derived hence authors concluded that littoral foraminifera only fed on particulate organic matter, a fourfold increase was observed after selected species of diatoms, chlorophytes and bacteria. Similarly, 2 weeks in the levels of diatom fatty acid biomarker inside the Duffield et al.(2014) observed a lack of positive response of foraminifera (Ward et al., 2003). In a field study of Schönfeld and foraminifera to net hauls dominated by dinoflagellates, except Numberger(2007), an increase in the populations of Elphidium in the case of the species Leptohalysis catella, which increased excavatum clavatum was found to occur simultaneously with the in abundance when dinoflagellates were provided as a food phytodetritus deposition. The authors suggested that Elphidium source. Alternatively to being a food source, dinoflagellate e. clavatum ingests fresh diatoms immediately upon deposition DNA occurrence in our specimens may be related to a from the water column and does not wait for incorporation of symbiotic relationship. Symbiosis between the dinoflagellates the organic detritus into the sediment. Our results support these and planktonic foraminifera is well-known (Gast and Caron, previous observations, and imply a predominantly planktivorous 1996; Garcia-Cuetos et al., 2005; Pochon and Gates, 2010; Siano feeding mode for Elphidium sp. and Haynesina sp. et al., 2010) but for benthic foraminifera only reported for large In addition to feeding, the acquisition of phototrophs by miliolids (Pawlowski et al., 2001). benthic foraminifera may be linked to photosymbionts or In some of our specimens (particularly in Elphidium sp. C3B the phenomenon of kleptoplasty, i.e., the assimilation and and M4D), there was high relative abundance of fungal DNA. maintenance of foreign chloroplasts. Both elphidiids and some The presence of fungal fruiting bodies of Ascomycetes has been nonioniids (e.g., Haynesina and Nonionellina) have the capacity observed before (Kohlmeyer, 1984, 1985) and it was suggested to retain chloroplasts from algal prey (e.g., Lopez, 1979; that the foraminiferal test chambers can serve as a protective Cedhagen, 1991; Pillet et al., 2011; Jauffrais et al., 2018). The niche for thin-walled fungal fruiting bodies (Kohlmeyer and active role of kleptoplasts in inorganic carbon assimilation Volkmann-Kohlmeyer, 1989) or that the protein-rich organic by H. germanica was recently demonstrated by a paired lining of the foraminiferal cell serves as nutrient source for the TEM-NanoSIMS observations in light conditions, suggesting a developing fungal ascocarps (Kohlmeyer, 1984). In our case, we functional photosynthetic role of kleptoplasts in H. germanica cannot be certain of the presence of active fungal parts within our (LeKieffre et al., 2018). In the same study, moderately 15N- specimens based on the presence of fungal DNA alone. It is also labeled kleptoplasts were observed in both light and darkness, possible that foraminifera acquired some fungal DNA attached which might indicate their involvement in nitrogen assimilation. onto sediment and diatom frustules while feeding. Kleptoplasts may be involved in carbon and nitrogen uptake The depth range, in which the specimens were found, was in other intertidal kleptoplast-bearing foraminiferal species as another significant factor for the observed intracellular eukaryote well, however, further analyses are needed to confirm their community variance inside our foraminiferal specimens function. Molecular analysis of the kleptoplasts of Haynesina (Figure 5). This makes sense, as sediment depth was also a sp. and Elphidium sp., have indicated that kleptoplasts in significant factor for the community variance in the sediments, these foraminifera originate exclusively from diatoms, however, meaning that different eukaryotes are found at different sediment there appears to be no clear specificity for diatom type (Pillet depths. Thus, foraminiferal specimens living at different et al., 2011). In photosymbiont-bearing foraminifera Pararotalia sediment depths would have access to different eukaryote calcariformata, the presence of 17 different endosymbiontic communities. The depth distribution of intertidal foraminifera diatoms has been recently linked to symbiont shuffling as in the sediment is typically focused on top sediments (e.g., an adaptation strategy under thermal stress (Schmidt et al., Langezaal et al., 2003; Thibault De Chanvalon et al., 2015), yet

Frontiers in Microbiology| www.frontiersin.org 12 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 13

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

intertidal foraminifera have been reported to occupy relatively be related to kleptoplasty. Moreover, our results suggest that irregular in-sediment distributions with living specimens the V9 region of the 18S rRNA gene can be used for occurring at tens of centimeters depth (Moodley and Hess, secure taxonomic assignment and phylogenetic placement of 1992). However, the activity of Ammonia sp. has been suggested foraminifera. Metabarcoding of the 18S V9 region allowed us to to decline and even enter a state of dormancy in low-oxygen confidently identify our specimens and assign their genetic types conditions (Maire et al., 2016; LeKieffre et al., 2017; Koho et al., (Elphidium sp. S5, Haynesina sp. S16, and Ammonia sp. T6). 2018) that typically prevail in deeper sediments. Based on our study, it is likely that some of the specimens living in deeper sediment horizons were still actively grazing in oxygenated AUTHOR CONTRIBUTIONS microenvironments, for example close to macrofaunal burrows. The sediments, especially at the Mokbaai site, were heavily P-MC carried out amplifications for MiSeq library preparations, bioturbated, which has been shown to be instrumental to the bioinformatics, and statistical analysis. IS extracted the DNA and vertical distribution of intertidal foraminifera (Bouchet et al., carried out initial tests with the primers. IS and KK designed 2009; Maire et al., 2016). and carried out sampling and processing of samples in the The general mechanisms of competition and adaptation in field. KK conceived the study and did the carbon and total different environmental conditions can generate and enhance the nitrogen analysis. CB assisted with the protocol for foraminiferal phenomenon of niche partitioning, which has been documented DNA extractions and with phylogenetic analysis, and did the among foraminifera (e.g., Aurahs et al., 2009; Weiner et al., genotyping. G-JR assisted with sampling co-ordination. P-MC, 2012). Benthic foraminifera are known to adapt to a variety IS, and KK contributed to interpretation of results and P-MC of habitats and this ability may be related and enhanced by drafted the manuscript. All authors contributed to the final their species-specific trophic preferences. It has been suggested version of the manuscript. that different feeding preferences among species could be an advantage in an environment where competition for space and food is high (Enge et al., 2014). This would be particularly FUNDING true in areas of high cell densities, which can be the case in intertidal microhabitats (e.g., Murray, 2006; Tsuchiya et al., This work was supported by the Academy of Finland (Project 2018). Moreover, intertidal zones are dynamic areas where Nos. 278827, 283453, and 312495). environmental conditions change rapidly, thus creating unique microhabitats. Therefore, a varied and species-specific trophic ACKNOWLEDGMENTS behavior, as suggested by our results, can be an advantage in such rapidly changing environments. However, future studies We are grateful to Lennart de Nooijer, Rick Hennekam and with more specimens are required to clarify the potential species- Sharyn Ossebaar from The Royal Netherlands Institute of Sea specific diet preferences of benthic foraminifera. Research (NIOZ) for making sampling on the island of Texel possible. We also appreciate the use of equipment and technical support from the Molecular Ecology and Systematics (MES) CONCLUSION laboratory at the University of Helsinki. The laboratory of DNA Sequencing and Genomics at the Institute of Biotechnology, To the best of our knowledge, this is the first study to use Helsinki Institute of Life Science (HiLIFE) carried out the MiSeq metabarcoding of the small subunit ribosomal DNA (SSU rDNA) sequencing and de-multiplexing of raw sequence reads. We with a view to gaining insights into the trophic preferences wish to acknowledge CSC – IT Center for Science, Finland, for of intertidal foraminifera and their role in the benthic food computational resources. web. In terms of their trophic behavior, benthic foraminifera are likely to have species-specific preferences. Ammonia sp. showed a tendency toward being a secondary consumer and SUPPLEMENTARY MATERIAL possibly preying actively on small eukaryote classes, such as Acoela, Nematoda, and Maxillopoda. Elphidiids and nonioniids The Supplementary Material for this article can be found (Elphidium sp., Haynesina sp.) showed a more herbivorous online at: https://www.frontiersin.org/articles/10.3389/fmicb. tendency with a clear preference for phototrophs, which could 2019.01169/full#supplementary-material

REFERENCES ribosomal RNA genes. PLoS One 4:e6372. doi: 10.1371/JOURNAL.PONE. 0006372 Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic Asano, K. (1951). “Illustrated Catalogue of Japanese tertiary smaller foraminifera,” local alignment search tool. J. Mol. Biol. 215, 403–410. doi: 10.1006/jmbi.1990. in Part 14. Rotaliidae, ed. L. W. Stach (Nuremberg: Hosokawa), 1–21. 9999 Aurahs, R., Grimm, G. W., Hemleben, V., Hemleben, C., and Kucera, M. (2009). Amaral-Zettler, L. A., McCliment, E. A., Ducklow, H. W., and Huse, S. M. Geographical distribution of cryptic genetic types in the planktonic foraminifer (2009). A method for studying protistan diversity using massively globigerinoides ruber. Mol. Ecol. 18, 1692–1706. doi: 10.1111/j.1365-294X.2009. parallel sequencing of V9 hypervariable regions of small-subunit 04136.x

Frontiers in Microbiology| www.frontiersin.org 13 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 14

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

Austin, H. A., Austin, W. E. N., and Paterson, D. M. (2005). Extracellular Ertan, K. T., Hemleben, V., and Hemleben, C. (2004). Molecular evolution of cracking and content removal of the benthic diatom Pleurosigma angulatum some selected benthic foraminifera as inferred from sequences of the small (Quekett) by the benthic foraminifera Haynesina germanica (Ehrenberg). Mar. subunit ribosomal DNA. Mar. Micropaleontol. 53, 367–388. doi: 10.1016/J. Micropaleontol. 57, 68–73. doi: 10.1016/J.MARMICRO.2005.07.002 MARMICRO.2004.08.001 Bé, A. W. H., Hemleben, C., Anderson, O. R., Spindler, M., Hacunda, J., Tuntivate- Esling, P., Lejzerowicz, F., and Pawlowski, J. (2015). Accurate multiplexing and Choy, S., et al. (1977). Laboratory and field observations of living planktonic filtering for high-throughput amplicon-sequencing. Nucleic Acids Res. 43, foraminifera. Micropaleontology 23, 155–179. doi: 10.2307/1485330 2513–2524. doi: 10.1093/nar/gkv107 Behnke, A., Engel, M., Christen, R., Nebel, M., Klein, R. R., and Stoeck, T. (2011). Garcia-Cuetos, L., Pochon, X., and Pawlowski, J. (2005). Molecular evidence for Depicting more accurate pictures of protistan community complexity using host–symbiont specificity in soritid foraminifera. Protist 156, 399–412. doi: pyrosequencing of hypervariable SSU rRNA gene regions. Environ. Microbiol. 10.1016/J.PROTIS.2005.08.003 13, 340–349. doi: 10.1111/j.1462-2920.2010.02332.x Gast, R. J., and Caron, D. A. (1996). Molecular phylogeny of symbiotic Bernhard, J. M. (2003). Potential symbionts in bathyal foraminifera. Science dinoflagellates from planktonic foraminifera and radiolaria. Mol. Biol. Evol. 13, 299:861. doi: 10.1126/science.1077314 1192–1197. doi: 10.1093/oxfordjournals.molbev.a025684 Bernhard, J. M., Tsuchiya, M., and Nomaki, H. (2018). Ultrastructural observations Goldstein, S. T. (1999). “Foraminifera: a biological overview,” in Modern on prokaryotic associates of benthic foraminifera: food, mutualistic symbionts, Foraminifera, ed. B. K. Sen Gupta (Dordrecht: Springer), 37–55. doi: 10.1007/ or parasites? Mar. Micropaleontol. 138, 33–45. doi: 10.1016/j.marmicro.2017. 0-306-48104-9_3 09.001 Guillou, L., Bachar, D., Audic, S., Bass, D., Berney, C., Bittner, L., et al. (2013). The Bird, C., Darling, K., Russell, A., Davis, C., Fehrenbacher, J., Free, A., et al. protist ribosomal reference database (PR2): a catalog of unicellular eukaryote (2017). Cyanobacterial endobionts within a major marine calcifier (Globigerina small sub-unit rRNA sequences with curated taxonomy. Nucleic Acids Res. 41, bulloides, Foraminifera) revealed by 16S rRNA metabarcoding. Biogeosciences D597–D604. doi: 10.1093/nar/gks1160 14, 901–920. doi: 10.5194/bg-14-901-2017 Hall, B. G. (2013). Building phylogenetic trees from molecular data with MEGA. Bird, C., Darling, K. F., Russell, A. D., Fehrenbacher, J. S., Davis, C. V., Free, A., et al. Mol. Biol. Evol. 30, 1229–1235. doi: 10.1093/molbev/mst012 (2018). 16S rRNA gene metabarcoding and TEM reveals different ecological Hallock, P. (2000). Symbiont-bearing foraminifera: harbingers of global change? strategies within the genus Neogloboquadrina (planktonic foraminifer). PLoS Micropaleontology 46, 95–104. doi: 10.2307/1486183 One 13:e0191653. doi: 10.1371/journal.pone.0191653 Hayward, B. W., Holzmann, M., Grenfell, H. R., Pawlowski, J., and Triggs, Bird, C., Schweizer, M., Roberts, A., Austin, W. E. N., Knudsen, K. L., Evans, C. M. (2004). Morphological distinction of molecular types in Ammonia – K. M., et al. (2019). The genetic diversity, morphology, biogeography, and towards a taxonomic revision of the world’s most commonly misidentified taxonomic designations of Ammonia (Foraminifera) in the Northeast Atlantic. foraminifera. Mar. Micropaleontol. 50, 237–271. doi: 10.1016/s0377-8398(03)00 Mar. Micropaleontol. (in press). doi: 10.1016/j.marmicro.2019.02.001 074-4 Boltovskoy, E., and Wright, R. (1976). Recent Foraminifera, ed. W. Junk The Holzmann, M., and Pawlowski, J. (1996). Preservation of foraminifera for DNA (Hague: Cambridge University Press). doi: 10.1017/S0016756800044332 extraction and PCR amplification. J. Foraminifer. Res. 26, 264–267. doi: 10.2113/ Bouchet, V. M. P., Sauriau, P.-G., Debenay, J.-P., Mermillod-Blondin, F., Schmidt, gsjfr.26.3.264 S., Amiard, J.-C., et al. (2009). Influence of the mode of macrofauna-mediated Holzmann, M., and Pawlowski, J. (2017). An updated classification of rotaliid bioturbation on the vertical distribution of living benthic foraminifera: first foraminifera based on ribosomal DNA phylogeny. Mar. Micropaleontol. 132, insight from axial tomodensitometry. J. Exp. Mar. Biol. Ecol. 371, 20–33. doi: 18–34. doi: 10.1016/J.MARMICRO.2017.04.002 10.1016/j.jembe.2008.12.012 Huson, D. H., Richter, D. C., Rausch, C., Dezulian, T., Franz, M., and Rupp, R. Caron, D. A., Countway, P. D., Savai, P., Gast, R. J., Schnetzer, A., Moorthi, S. D., (2007). Dendroscope: an interactive viewer for large phylogenetic trees. BMC et al. (2009). Defining DNA-based operational taxonomic units for microbial- Bioinformatics 8:460. doi: 10.1186/1471-2105-8-460 eukaryote ecology. Appl. Environ. Microbiol. 75, 5797–5808. doi: 10.1128/AEM. Jauffrais, T., LeKieffre, C., Koho, K. A., Tsuchiya, M., Schweizer, M., Bernhard, 00298-09 J. M., et al. (2018). Ultrastructure and distribution of kleptoplasts in benthic Cedhagen, T. (1991). Retention of chloroplasts and bathymetric distribution in foraminifera from shallow-water (photic) habitats. Mar. Micropaleontol. 138, the sublittoral foraminiferan Nonionellina labradorica. Ophelia 33, 17–30. doi: 46–62. doi: 10.1016/j.marmicro.2017.10.003 10.1080/00785326.1991.10429739 Jeffreys, R. M., Fisher, E. H., Gooday, A. J., Larkin, K. E., Billett, D. S. M., and Darling, K. F., Schweizer, M., Knudsen, K. L., Evans, K. M., Bird, C., Roberts, Wolff, G. A. (2015). The trophic and metabolic pathways of foraminifera A., et al. (2016). The genetic diversity, phylogeography and morphology of in the Arabian Sea: evidence from cellular stable isotopes. Biogeosciences 12, Elphidiidae (Foraminifera) in the Northeast Atlantic. Mar. Micropaleontol. 129, 1781–1797. doi: 10.5194/bg-12-1781-2015 1–23. doi: 10.1016/j.marmicro.2016.09.001 Kohlmeyer, J. (1984). Tropical marine fungi. Mar. Ecol. 5, 329–378. doi: 10.1111/j. Dasilva, C. R., Li, W. K. W., and Lovejoy, C. (2014). Phylogenetic diversity 1439-0485.1984.tb00130.x of eukaryotic marine microbial plankton on the Scotian Shelf Northwestern Kohlmeyer, J. (1985). Marine fungi (Ascomycetes) within and on tests of Atlantic Ocean. J. Plankton Res. 36, 344–363. doi: 10.1093/plankt/fbt123 Foraminifera. Mar. Biol. 90, 147–149. doi: 10.1007/BF00428226 de Vargas, C., Audic, S., Henry, N., Decelle, J., Mahé, F., Logares, R., et al. (2015). Kohlmeyer, J., and Volkmann-Kohlmeyer, B. (1989). Hawaiian marine fungi, Ocean plankton. Eukaryotic plankton diversity in the sunlit ocean. Science including two new genera of Ascomycotina. Mycol. Res. 92, 410–421. doi: 10. 348:1261605. doi: 10.1126/science.1261605 1016/s0953-7562(89)80185-6 Duffield, C. J., Edvardsen, B., Eikrem, W., and Alve, E. (2014). Effects of different Koho, K. A., LeKieffre, C., Nomaki, H., Salonen, I., Geslin, E., Mabilleau, G., potential food sources on upper-bathyal benthic foraminifera: an experiment et al. (2018). Changes in ultrastructural features of the foraminifera Ammonia with propagules. J. Foraminifer. Res. 44, 416–433. doi: 10.2113/gsjfr.44.4.416 spp. in response to anoxic conditions: field and laboratory observations. Dupuy, C., Rossignol, L., Geslin, E., and Pascal, P.-Y. (2010). Predation of mudflat Mar. Micropaleontol. 138, 72–82. doi: 10.1016/J.MARMICRO.2017. meio-macrofaunal metazoans by a calcareous foraminifer, Ammonia tepida 10.011 (Cushman, 1926). J. Foraminifer. Res. 40, 305–312. doi: 10.2113/gsjfr.40.4.305 Koho, K. A., Piña-Ochoa, E., Geslin, E., and Risgaard-Petersen, N. (2011). Edgar, R. C. (2004). MUSCLE: a multiple sequence alignment method with reduced Vertical migration, nitrate uptake and denitrification: survival mechanisms time and space complexity. BMC Bioinformatics 5:113. doi: 10.1186/1471-2105- of foraminifers (Globobulimina turgida) under low oxygen conditions. 5-113 FEMS Microbiol. Ecol. 75, 273–283. doi: 10.1111/j.1574-6941.2010. Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C., and Knight, R. (2011). 01010.x UCHIME improves sensitivity and speed of chimera detection. Bioinformatics Kumar, S., Stecher, G., and Tamura, K. (2016). MEGA7: molecular evolutionary 27, 2194–2200. doi: 10.1093/bioinformatics/btr381 genetics analysis version 7.0 for bigger datasets. Mol. Biol. Evol. 33, 1870–1874. Enge, A. J., Witte, U., Kucera, M., and Heinz, P. (2014). Uptake of phytodetritus doi: 10.1093/molbev/msw054 by benthic foraminifera under oxygen depletion at the Indian margin (Arabian Langer, M. R., and Bell, C. J. (1995). Toxic foraminifera: innocent until proven Sea). Biogeosciences 11, 2017–2026. doi: 10.5194/bg-11-2017-2014 guilty. Mar. Micropaleontol. 24, 205–214. doi: 10.1016/0377-8398(94)00023-G

Frontiers in Microbiology| www.frontiersin.org 14 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 15

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

Langer, M. R., and Leppig, U. (2000). Molecular phylogenetic status of Ammonia Morard, R., Garet-Delmas, M.-J., Mahé, F., Romac, S., Poulain, J., Kucera, M., et al. catesbyana (D’Orbigny 1839), an intertidal foraminifer from the North Sea. (2018). Surface ocean metabarcoding confirms limited diversity in planktonic Neues Jahrb. Geol. Palaontol. Monatsh. 9, 545–556. foraminifera but reveals unknown hyper-abundant lineages. Sci. Rep. 8:2539. Langezaal, A., Ernst, S., Haese, R., van Bergen, P., and van der Zwaan, G. (2003). doi: 10.1038/s41598-018-20833-z Disturbance of intertidal sediments: the response of bacteria and foraminifera. Murray, J. W. (2006). Ecology and Applications of Benthic Foraminifera. Cambridge: Estuar. Coast. Shelf Sci. 58, 249–264. doi: 10.1016/s0272-7714(03)00078-7 Cambridge University Press. doi: 10.1017/CBO9780511535529 Larkin, K. E., Gooday, A. J., Woulds, C., Jeffreys, R. M., Schwartz, M., Cowie, G., Nomaki, H., Heinz, P., Nakatsuka, T., Shimanaga, M., and Kitazato, H. (2005). et al. (2014). Uptake of algal carbon and the likely synthesis of an "essential" fatty Species-specific ingestion of organic carbon by deep-sea benthic foraminifera acid by Uvigerina ex. gr. semiornata (Foraminifera) within the Pakistan margin and meiobenthos: in situ tracer experiments. Limnol. Oceanogr. 50, 134–146. oxygen minimum zone: evidence from fatty acid biomarker and 13C tracer doi: 10.4319/lo.2005.50.1.0134 experiments. Biogeosciences 11, 3729–3738. doi: 10.5194/bg-11-3729-2014 Nomaki, H., Heinz, P., Nakatsuka, T., Shimanaga, M., Ohkouchi, N., Ogawa, N., Lecroq, B., Lejzerowicz, F., Bachar, D., Christen, R., Esling, P., Baerlocher, L., et al. et al. (2006). Different ingestion patterns of 13C-labeled bacteria and algae by (2011). Ultra-deep sequencing of foraminiferal microbarcodes unveils hidden deep-sea benthic foraminifera. Mar. Ecol. Prog. Ser. 310, 95–108. doi: 10.3354/ richness of early monothalamous lineages in deep-sea sediments. Proc. Natl. meps310095 Acad. Sci. U.S.A. 108, 13177–13182. doi: 10.1073/pnas.1018426108 Nomura, R. (2003). Assessing the roles of artificial vs. natural impacts on Lee, J. J., McEnery, M., Pierce, S., Freudenthal, H. D., and Muller, W. A. (1966). brackish lake environments?: foraminiferal evidence from Lake Nakaumi, Tracer experiments in feeding littoral foraminifera. J. Protozool. 13, 659–670. southwest Japan. J. Geol. Soc. Japan 109, 197–214. doi: 10.5575/geosoc. doi: 10.1111/j.1550-7408.1966.tb01978.x 109.197 Lei, Y., Li, T., Nigam, R., Holzmann, M., and Lyu, M. (2017). Environmental Nomura, R. and Seto, K. (1992). “Benthic foraminifera from brackish Lake significance of morphological variations in the foraminifer Ammonia Nakanoumi, San-in district, southwestern Honshu, Japan,” in Centenary aomoriensis (Asano, 1951) and its molecular identification: a study from Japanese Micropaleontology, eds K. Ishizaki and T. Saito (Tokyo: Terra Scientific the Yellow Sea and East China Sea, PR China. Palaeogeogr. Palaeoclimatol. Publishing Company), 227–240. Palaeoecol. 483, 49–57. doi: 10.1016/j.palaeo.2016.05.010 Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’Hara, R. B., LeKieffre, C., Jauffrais, T., Geslin, E., Jesus, B., Bernhard, J. M., Giovani, M.-E., et al. et al. (2015). vegan: Community Ecology Package. R Package version 2.3-1. (2018). Inorganic carbon and nitrogen assimilation in cellular compartments of Pawlowski, J. (2000). Introduction to the molecular systematics of foraminifera. a benthic kleptoplastic foraminifer. Sci. Rep. 8:10140. doi: 10.1038/s41598-018- Micropaleontology 46, 1–12. doi: 10.2307/1486176 28455-1 Pawlowski, J., Christen, R., Lecroq, B., Bachar, D., Shahbazkia, H. R., Amaral- LeKieffre, C., Spangenberg, J. E., Mabilleau, G., Escrig, S., Meibom, A., and Geslin, Zettler, L., et al. (2011). Eukaryotic richness in the abyss: insights from pyrotag E. (2017). Surviving anoxia in marine sediments: the metabolic response of sequencing. PLoS One 6:e18169. doi: 10.1371/journal.pone.0018169 ubiquitous benthic foraminifera (Ammonia tepida). PLoS One 12:e0177604. Pawlowski, J., and Holzmann, M. (2008). Diversity and geographic distribution of doi: 10.1371/journal.pone.0177604 benthic foraminifera: a molecular perspective. Biodivers. Conserv. 17, 317–328. Lipps, J. H. (1983). “Biotic interactions in benthic foraminifera,” in Biotic doi: 10.1007/s10531-007-9253-8 Interactions in Recent and Fossil Benthic Communities. Topics in Geobiology, Pawlowski, J., and Holzmann, M. (2014). A plea for DNA barcoding of eds M. J. S. Tevesz and P. L. McCall (Boston, MA: Springer), 331–376. doi: foraminifera. J. Foraminifer. Res. 44, 62–67. doi: 10.2113/gsjfr.44.1.62 10.1007/978-1-4757-0740-3_8 Pawlowski, J., Holzmann, M., Fahrni, J. F., and Hallock, P. (2001). Molecular Logares, R., Sunagawa, S., Salazar, G., Cornejo-Castillo, F. M., Ferrera, I., Sarmento, identification of algal endosymbionts in large miliolid foraminifera: 1. H., et al. (2014). Metagenomic 16S rDNA Illumina tags are a powerful Chlorophytes. J. Eukaryot. Microbiol. 48, 362–367. doi: 10.1111/j.1550-7408. alternative to amplicon sequencing to explore diversity and structure of 2001.tb00325.x microbial communities. Environ. Microbiol. 16, 2659–2671. doi: 10.1111/1462- Pawluczyk, M., Weiss, J., Links, M. G., Egaña Aranguren, M., Wilkinson, 2920.12250 M. D., and Egea-Cortines, M. (2015). Quantitative evaluation of bias in PCR Lopez, E. (1979). Algal chloroplasts in the protoplasm of three species of benthic amplification and next-generation sequencing derived from metabarcoding foraminifera: taxonomic affinity, viability and persistence. Mar. Biol. 53, 201– samples. Anal. Bioanal. Chem. 407, 1841–1848. doi: 10.1007/s00216-014- 211. doi: 10.1007/BF00952427 8435-y Maire, O., Barras, C., Gestin, T., Nardelli, M., Romero-Ramirez, A., Duchêne, Pillet, L., de Vargas, C., and Pawlowski, J. (2011). Molecular Identification of J., et al. (2016). How does macrofaunal bioturbation influence the vertical sequestered diatom chloroplasts and kleptoplastidy in foraminifera. Protist 162, distribution of living benthic foraminifera? Mar. Ecol. Prog. Ser. 561, 83–97. 394–404. doi: 10.1016/J.PROTIS.2010.10.001 doi: 10.3354/meps11929 Pillet, L., Fontaine, D., and Pawlowski, J. (2012). Intra-genomic ribosomal RNA McMurdie, P. J., and Holmes, S. (2013). phyloseq: an R package for reproducible polymorphism and morphological variation in Elphidium macellum suggests interactive analysis and graphics of microbiome census data. PLoS One inter-specific hybridization in foraminifera. PLoS One 7:e32373. doi: 10.1371/ 8:e61217. doi: 10.1371/journal.pone.0061217 journal.pone.0032373 Miller, A. A. L., Scott, D. B., and Medioli, F. S. (1982). Elphidium excavatum Pillet, L., Voltski, I., Korsun, S., and Pawlowski, J. (2013). Molecular phylogeny (Terquem); ecophenotypic versus subspecific variation. J. Foraminifer. Res. 12, of Elphidiidae (foraminifera). Mar. Micropaleontol. 103, 1–14. doi: 10.1016/J. 116–144. doi: 10.2113/gsjfr.12.2.116 MARMICRO.2013.07.001 Mojtahid, M., Geslin, E., Coynel, A., Gorse, L., Vella, C., Davranche, A., et al. Pitsch, G., Bruni, E. P., Forster, D., Qu, Z., Sonntag, B., Stoeck, T., et al. (2019). (2016). Spatial distribution of living (rose bengal stained) benthic foraminifera Seasonality of Planktonic Freshwater Ciliates: are analyses based on V9 regions in the Loire estuary (western France). J. Sea Res. 118, 1–16. doi: 10.1016/j.seares. of the 18S rRNA gene correlated with morphospecies counts? Front. Microbiol. 2016.02.003 10:248. doi: 10.3389/fmicb.2019.00248 Moodley, L., Boschker, H., Middelburg, J., Pel, R., Herman, P., de Deckere, E., Pochon, X., and Gates, R. D. (2010). A new Symbiodinium clade (Dinophyceae) et al. (2000). Ecological significance of benthic foraminifera: 13C labelling from soritid foraminifera in Hawai’i. Mol. Phylogenet. Evol. 56, 492–497. doi: experiments. Mar. Ecol. Prog. Ser. 202, 289–295. doi: 10.3354/meps202289 10.1016/j.ympev.2010.03.040 Moodley, L., and Hess, C. (1992). Tolerance of infaunal benthic foraminifera Ray, J. L., Althammer, J., Skaar, K. S., Simonelli, P., Larsen, A., Stoecker, D., et al. for low and high oxygen concentrations. Biol. Bull. 183, 94–98. doi: 10.2307/ (2016). Metabarcoding and metabolome analyses of copepod grazing reveal 1542410 feeding preference and linkage to metabolite classes in dynamic microbial Morard, R., Darling, K. F., Mahé, F., Audic, S., Ujiié, Y., Weiner, A. K. M., plankton communities. Mol. Ecol. 25, 5585–5602. doi: 10.1111/mec.13844 et al. (2015). PFR 2?: a curated database of planktonic foraminifera 18S Roberts, A., Austin, W., Evans, K., Bird, C., Schweizer, M., and Darling, K. ribosomal DNA as a resource for studies of plankton ecology, biogeography (2016). A new integrated approach to taxonomy: the fusion of molecular and and evolution. Mol. Ecol. Resour. 15, 1472–1485. doi: 10.1111/1755-0998. morphological systematics with type material in benthic foraminifera. PLoS One 12410 11:e0158754. doi: 10.1371/journal.pone.0158754

Frontiers in Microbiology| www.frontiersin.org 15 May 2019| Volume 10| Article 1169 fmicb-10-01169 May 28, 2019 Time: 14:28 # 16

Chronopoulou et al. Trophic Behavior and Identity of Benthic Foraminifera

Salava, A., Aho, V., Lybeck, E., Pereira, P., Paulin, L., Nupponen, I., et al. (2017). Thibault De Chanvalon, A., Metzger, E., Mouret, A., Cesbron, F., Knoery, J., Loss of cutaneous microbial diversity during first 3 weeks of life in very low Rozuel, E., et al. (2015). Two-dimensional distribution of living benthic birthweight infants. Exp. Dermatol. 26, 861–867. doi: 10.1111/exd.13312 foraminifera in anoxic sediment layers of an estuarine mudflat (Loire Sawaya, N. A., Djurhuus, A., Closek, C. J., Hepner, M., Olesin, E., Visser, L., Estuary, France). Biogeosci. Discuss. 12, 6219–6234. doi: 10.5194/bg-12- et al. (2019). Assessing eukaryotic biodiversity in the Florida Keys National 6219-2015 Marine Sanctuary through environmental DNA metabarcoding. Ecol. Evol. 9, Tsuchiya, M., Chikaraishi, Y., Nomaki, H., Sasaki, Y., Tame, A., Uematsu, K., et al. 1029–1040. doi: 10.1002/ece3.4742 (2018). Compound-specific isotope analysis of benthic foraminifer amino acids Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M., Hollister, suggests microhabitat variability in rocky-shore environments. Ecol. Evol. 8, E. B., et al. (2009). Introducing mothur: open-source, platform-independent, 8380–8395. doi: 10.1002/ece3.4358 community-supported software for describing and comparing microbial Tsuchiya, M., Toyofuku, T., Uematsu, K., Brüchert, V., Collen, J., Yamamoto, H., communities. Appl. Environ. Microbiol. 75, 7537–7541. doi: 10.1128/AEM. et al. (2015). Cytologic and genetic characteristics of endobiotic bacteria and 01541-09 kleptoplasts of Virgulinella fragilis (Foraminifera). J. Eukaryot. Microbiol. 62, Schmidt, C., Morard, R., Romero, O., and Kucera, M. (2018). Diverse 454–469. doi: 10.1111/jeu.12200 internal symbiont community in the endosymbiotic foraminifera Pararotalia Ward, J. N., Pond, D. W., and Murray, J. W. (2003). Feeding of benthic foraminifera calcariformata: implications for symbiont shuffling under uhermal stress. Front. on diatoms and sewage-derived organic matter: an experimental application Microbiol. 9:2018. doi: 10.3389/fmicb.2018.02018 of lipid biomarker techniques. Mar. Environ. Res. 56, 515–530. doi: 10.1016/ Schönfeld, J., and Numberger, L. (2007). The benthic foraminiferal response to the s0141-1136(03)00040-0 2004 spring bloom in the western Baltic Sea. Mar. Micropaleontol. 65, 78–95. Weiner, A., Aurahs, R., Kurasawa, A., Kitazato, H., and Kucera, M. (2012). Vertical doi: 10.1016/j.marmicro.2007.06.003 niche partitioning between cryptic sibling species of a cosmopolitan marine Schuelke, T., Pereira, T. J., Hardy, S. M., and Bik, H. M. (2018). Nematode- planktonic protist. Mol. Ecol. 21, 4063–4073. doi: 10.1111/j.1365-294X.2012. associated microbial taxa do not correlate with host phylogeny, geographic 05686.x region or feeding morphology in marine sediment habitats. Mol. Ecol. 27, Woulds, C., Cowie, G. L., Levin, L. A., Andersson, J. H., Middelburg, 1930–1951. doi: 10.1111/mec.14539 J. J., Vandewiele, S., et al. (2007). Oxygen as a control on sea Schweizer, M., Pawlowski, J., Kouwenhoven, T. J., Guiard, J., and van der Zwaan, floor biological communities and their roles in sedimentary carbon B. (2008). Molecular phylogeny of Rotaliida (Foraminifera) based on complete cycling. Limnol. Oceanogr. 52, 1698–1709. doi: 10.4319/lo.2007.52. small subunit rDNA sequences. Mar. Micropaleontol. 66, 233–246. doi: 10.1016/ 4.1698 J.MARMICRO.2007.10.003 Xiang, R., Yang, Z., Saito, Y., Fan, D., Chen, M., Guo, Z., et al. (2008). Schweizer, M., Polovodova, I., Nikulina, A., and Schönfeld, J. (2011). Molecular Paleoenvironmental changes during the last 8400 years in the southern identification of Ammonia and Elphidium species (Foraminifera, Rotaliida) Yellow Sea: benthic foraminiferal and stable isotopic evidence. from the Kiel Fjord (SW Baltic Sea) with rDNA sequences. Helgol. Mar. Res. Mar. Micropaleontol. 67, 104–119. doi: 10.1016/J.MARMICRO.2007. 65, 1–10. doi: 10.1007/s10152-010-0194-3 11.002 Siano, R., Montresor, M., Probert, I., Not, F., and de Vargas, C. (2010). Pelagodinium gen. nov. and P. béii comb. nov., a dinoflagellate symbiont of Conflict of Interest Statement: The authors declare that the research was planktonic foraminifera. Protist 161, 385–399. doi: 10.1016/j.protis.2010.01.002 conducted in the absence of any commercial or financial relationships that could Somervuo, P., Koskinen, P., Mei, P., Holm, L., Auvinen, P., and Paulin, L. (2018). be construed as a potential conflict of interest. BARCOSEL: a tool for selecting an optimal barcode set for high-throughput sequencing. BMC Bioinformatics 19:257. doi: 10.1186/s12859-018-2262-7 Copyright © 2019 Chronopoulou, Salonen, Bird, Reichart and Koho. This is an Suhr, S., Alexander, S., Gooday, A., Pond, D., and Bowser, S. (2008). Trophic modes open-access article distributed under the terms of the Creative Commons Attribution of large Antarctic Foraminifera: roles of carnivory, omnivory, and detritivory. License (CC BY). The use, distribution or reproduction in other forums is permitted, Mar. Ecol. Prog. Ser. 371, 155–164. doi: 10.3354/meps07693 provided the original author(s) and the copyright owner(s) are credited and that the Takata, H., Takayasu, K., and Hasegawa, S. (2006). Foraminifera in an organic-rich, original publication in this journal is cited, in accordance with accepted academic brackish-water lagoon, lake Saroma, Hokkaido, Japan. J. Foraminifer. Res. 36, practice. No use, distribution or reproduction is permitted which does not comply 44–60. doi: 10.2113/36.1.44 with these terms.

Frontiers in Microbiology| www.frontiersin.org 16 May 2019| Volume 10| Article 1169